[FFmpeg-devel] [PATCH] ARM: NEON optimised simple_idct

[Måns Rullgård]

M?ns Rullg?rd <mans at mansr.com> writes:

> Michael Niedermayer <michaelni at gmx.at> writes:
>
>> Still its likely better to use a transposed permutation instead of
>> the identity one as this means 1 transpose less in a SIMD IDCT.
>
> That idea struck me as well.  I'll try it out.

It saves 8 instructions per block, it seems to be a slightly faster
too, so I've made the changed in my tree.

-- 
M?ns Rullg?rd
mans at mansr.com